INDEX

Explanations

references to specific entities or subjects, particularly those labeled with a "J" designation

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 26-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L26R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.26.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

ulia

-0.15

HTTPHeader

-0.15

ames

-0.15

ohn

-0.15

_cmos

-0.15

ERNEL

-0.15

ulan

-0.15

 trai

-0.15

avic

-0.14

clide

-0.14

POSITIVE LOGITS

abil

0.17

ume

0.17

umpt

0.15

bol

0.15

ispecies

0.15

edula

0.15

 pretext

0.14

NK

0.14

amedi

0.14

Activations Density 0.033%

references to specific entities or subjects, particularly those labeled with a "J" designation

No Comments

No Known Activations